The goal of the proposed research is to study how nonsense codons reduce mRNA abundance in mammalian cells. Nonsense codons can arise as the consequence of correct or incorrect RNA editing, nonproductive DNA rearrangements, the mutation of germline or somatic DNA, the inefficient or inaccurate removal of introns, or the failure to frameshift. We will continue to characterize the effect of nonsense codons within the gene for human triosephosphate isomerase (TPI). Most studies will involve the transient co-transfection of mouse cells with a test TPI allele, which will be driven by a promoter that suits the purpose of the experiment, and a reference allele, which will control for variations in transfection efficiency and RNA recovery. Using this approach and the transiently inducible c-fos promoter to drive TPI gene expression, we have shown that nonsense codons within the first three-quarters of the translated portion of TPI mRNA mediate the decay of nucleus-associated but not cytoplasmic mRNA. Translating ribosomes are implicated in the decay process from our findings that the nonsense-mediated reduction of TPI mRNA abundance is abrogated by either (i) a hairpin structure in the 5' untranslated region that acts in cis to inhibit translation initiation or (ii) a mutant tRNA that acts in trans to suppress the nonsense codon. Translating ribosomes are known to exist only in the cytoplasm. Therefore, the decay of nucleus- associated nonsense mRNA may be mediated by cytoplasmic ribosomes that translate newly synthesized mRNA as it is being exported from the nucleus to the cytoplasm. One aim will be to determine if mRNA export is required for nonsense mRNA decay. The export of nonsense TPI mRNA will be (i) made dependent on the Rev protein of human immunodeficiency virus-1 using L cells or (ii) blocked by RL- l a monoclonal antibody to the nuclear pore complex, using X laevis oocytes. Another aim will be to determine if the mechanism of nonsense mRNA decay in mammalian cells involves a 5'->3' exonuclease. as it does in S. cerevisiae. To this end, structural barriers to exoribonucleolytic activity will be generated at various positions within nonsense TPI mRNA in order to facilitate the detection of decay intermediates. The controversial issue of whether or not nonsense codons can influence splice site choice will be addressed. Additionally, cis- acting sequences of nonsense mRNA decay will be identified. Since studies of mammalian mRNAs other than TPI mRNA also indicate that nonsense codons can affect the metabolism of nucleus-associated but not cytoplasmic RNA, we will continue to supplement our studies of TPI mRNA with studies of other mRNAs in order to establish mechanistic similarities in the decay of distinct nonsense mRNAs.